Remove PDF metadata

What metadata does a PDF contain, and what can it give away?

A PDF normally carries eight standard fields: Title, Author, Subject, Keywords, Creator, Producer, CreationDate and ModDate, and often an XMP packet that repeats them. Word, Google Docs, scanners and design apps fill these in silently on export. The Author field often picks up the name on the computer account, which can be a full name, a nickname or a colleague's name if the file started from their template.

Creator and Producer reveal the exact software and version, and the dates show when a draft was really made, which can contradict the date typed on the page. Journalists, lawyers, job seekers and anyone publishing documents online have good reasons to check this before sending. Search engines index PDF titles too, so a leftover title like "Copy of draft v7 FINAL" can end up in search results.

Which metadata fields does Remove all delete?

Remove all deletes every entry in the Info dictionary, standard and custom, plus the document XMP packet, any page-level XMP packets and the application-private PieceInfo data that Adobe apps store. After saving, the tool re-opens the output and fails with an error rather than a download if any Info field or the XMP packet is still present, so the before and after table always reflects the real file.

Edit fields is selective. Title, Author, Subject, Keywords, Creator and Producer can each be rewritten, and a field left blank is deleted. Dates and custom entries are removed only when their checkbox is ticked, and the XMP option is ticked by default whenever a packet exists. The file is saved without adding a new producer name or modification date, so nothing new is written in place of what was removed.

Where is PDF metadata stored?

PDF metadata is stored in two places: the Info dictionary and an XMP packet. The Info dictionary holds the familiar fields shown in a reader's Properties window. XMP, Adobe's Extensible Metadata Platform, is a separate XML packet attached to the document, and sometimes to individual pages, that usually repeats the same details and may add editing history and document IDs. PDF 2.0, published as ISO 32000-2, treats XMP as the primary location and deprecates most of the Info dictionary.

Clearing only the Properties fields therefore leaves a second copy behind, and many simple editors do exactly that. A recipient who opens the file in a text editor or a metadata viewer still sees the old author and software in the XML. This is why Remove all deletes both locations, and why the XMP option should stay ticked when editing fields, so old values don't linger.

Should I remove all metadata or edit the fields?

Remove all is the safer choice for anonymous sharing, such as a whistleblower document, a blind peer review or a public tender response, because it leaves no field to overlook. Edit fields is better when the document should look tidy and still carry a sensible name, for instance a brochure or report published on a website:

• Set a clear Title so readers and search engines show a useful name.
• Replace a personal Author with a company or team name.
• Clear Keywords and Subject left over from a template.
• Remove creation and modification dates if timing is sensitive.

What does metadata removal not cover?

It does not touch anything inside the page content. Names in headers, comments, form fields or tracked changes baked into the PDF are not metadata; use Redact PDF to remove those. Password-protected files are opened with their password and saved without it, so run Protect PDF afterwards if the cleaned copy needs one.

Embedded photos can carry their own camera data such as GPS location, which this tool doesn't rewrite, and attached files keep their own properties. For those, export the images again without location data before building the PDF. A digitally signed PDF loses its valid signature when it is rewritten, so clean the metadata before the document is signed rather than after.